Higher intake of plant food is associated with lower risk of cancer in humans. There are a large number of bioactive constituents present in plant foods or derived from them in the human GI tract. There is an extensive body of animal experimental literature to show that many of these bioactive compounds have specific effects at the biochemical and molecular level - including important influences on xenobiotic metabolism. At present, however, evidence regarding mechanisms of cancer risk reduction in humans is almost totally lacking. To test the hypothesis that induction of Phase II enzyme systems through increased intakes of vegetable and fruits is a plausible mechanism for reduced cancer risk the applicants propose to examine the effects of vegetable and fruit consumption on glutathione-S-transferase (GST) isoenzyme activity in humans. The specific aims of the proposal are: 1) to examine, in a cross-sectional design, the associations of vegetable and fruit intake with serum GST alpha concentrations, overall serum GST activity and lymphocyte GSTu activity; 2) to determine whether these associations differ by GSTM1 genotype; 3) to measure the effect of feeding specific vegetables and fruit under controlled dietary conditions on serum GST alpha concentration, overall serum GST activity and lymphocyte GSTu activity; and 4) to explore whether the results are modulated by other xenobiotic metabolizing enzymes. The proposed project will be implemented in two parts: 1) a cross-sectional study and 2) a controlled randomized cross-over feeding trial. For the cross-sectional study, the applicants will recruit 200 non-smokers, 20-40 years of age. They will genotype them for GST1, measure serum GST alpha concentration, overall serum GST activity and lymphocyte GSTu activity, and will assess diet by food frequency questionnaire, 3-day food records, and the following serum and plasma markers of dietary exposure: ascorbic acid, beta carotene, thiocyanate and phytoestrogens (genistein, daidzein, equol, O-desmethylangolensin, enterolactone and enterdiol). They will correlate the measures of GST in the GSTM1 null and non-null participants with consumption of vegetables and fruit, nutrient intakes and the serum markers. For the feeding trial they will recruit a subset of 48 participants (24 GSTM1 null and 24 non-null) from the 200 in the cross-sectional study to take part in four 7-day feeding periods to examine the effects of various vegetable groupings (cruciferous vegetables, allium vegetables, and a mixture of vegetables and fruits) compared to a basal diet on induction of GST activity under controlled dietary conditions. They will measure serum GSTalpha concentration, overall serum GST activity and lymphocyte GSTu activity at the end of each feeding period. Concurrently, they will monitor induction of Phase I enzyme CYP1A2 by measuring urinary caffeine metabolites to determine whether alterations in GST activity are specific Phase II effects or reflect overall shifts in xenobiotic metabolism.